Hi Bill,
Let me clarify where we disagree. You believe that
"the "type" field of the Content-type header would become
useless, because it could not be used in such a manner as
to present accurate information as to the viewing resources
necessary for the document being sent"
(this is out of context, but I hope without distorting your position).
I don't believe that the specification of necessary viewing resources is
the only purpose of the "type" field, and I find the other purposes useful.
Specifying that the body part is intended for use with an application is
useful to me, in and of itself, regardless of the viewing resources involved.
Now, I can also see how a resource list could be useful. Here's a proposal
you'll probably hate: specify a Mixed-Media subtype of Application, and
include a resource list as one field. Then you can check the resources
against your supported list while I check the application against mine.
My viewpoint is no doubt warped by the fact that I am supporting a fax
gateway so my only "viewing resource" is a bitmap. It may be that for
most display-agent implementations your interpretation of the "type" field
is more accurate. (I see it reducing then to Text, Bitmap and Not-Intended.)
Any implementors wish to comment?
Jim